Every platform rep will tell you to make more video. Every creative agency will quote you for video. And plenty of accounts I audit are spending most of their production budget on video while their best-performing ad is a static someone made in an afternoon.
The honest answer is that neither format wins in general. They win in different situations, and the situation is usually decided by three things: what stage of the funnel the ad is doing work in, how much you can spend producing it, and how much your product needs to be explained versus shown.
What each format is structurally good at
Video buys attention and explains. It can demonstrate a product in use, establish a problem before offering a solution, carry a spoken argument, and hold someone for fifteen seconds who would have scrolled past an image in one. If the buyer needs to understand something before they can want it, video does work that a static cannot.
Statics buy frequency and clarity. They are cheap enough to produce in volume, which means you can run twelve variations instead of two. They load instantly, they work with the sound off and the screen small, and they deliver a single message without asking for time. If the buyer already understands the category and the job is to state an offer, a static often outperforms the video that cost twenty times more.
That last point is the one people resist. A well-designed static is not a downgrade from video. It is a different instrument.
The funnel-stage rule
This is the fastest way to make the decision, and it holds up across most accounts I have worked on:
Cold, category-unaware audiences → video. The person does not know they have the problem, or does not know a solution like yours exists. You need to establish context, and context takes time. A static asking a stranger to buy something they have never heard of is asking too much.
Warm, category-aware audiences → statics, usually. The person knows what the product is. Now the job is the offer, the differentiator, the proof, the objection. Those are sentences and images, not scenes. This is where statics quietly outperform and where most accounts are over-investing in video.
Retargeting → statics, almost always. They have already seen the video. They do not need the explanation again; they need a reason to act now. Shipping, guarantee, price, social proof, urgency. Cheap, specific, in volume.
The exception that breaks the rule: visually-driven products. Fashion, food, interiors, travel, anything where the product is the appeal. There, video and rich imagery both work at every stage, because seeing it is the argument.
The cost side that decides more than performance does
Performance comparisons usually ignore the thing that matters most in practice: you do not have unlimited production capacity.
A decent static costs a designer an hour or two. A decent video costs a shoot, an edit, and a round of revisions — days, and often an external vendor. Which means that in the same month, at the same budget, you can test fifteen statics or two videos.
That ratio is the real argument. Creative testing is a volume game, and the format that lets you run more tests usually wins on learning rate even when it loses on any single head-to-head. See creative testing on paid social for how to run that properly.
The practical consequence: statics are how you find the message. Video is how you scale the message once you know what it is. Making expensive video before you know what to say is the most common way to waste a creative budget.
Reading the metrics honestly
Comparing formats using the wrong metric produces confident nonsense.
Video gets better engagement metrics almost by definition. Thruplays, view rates, watch time — statics cannot produce these at all. Comparing a video’s engagement to a static’s is not a comparison.
Statics often get better CTR. Also structural: a static asks for the click immediately, whereas a video asks for attention first. Higher CTR does not mean the static is better; it means it is doing a different job.
Judge both on cost per outcome, at the account level. Cost per purchase, cost per qualified lead, or — better — what happened to blended efficiency when you shifted the mix. Not CTR, not CPM, not view rate. This is the same discipline as reading ROAS properly: platform-native metrics flatter whatever the platform wants more of.
One more that catches people: video usually has a lower CPM and a higher cost per click. Cheap impressions are not the goal. It is entirely possible to buy a lot of very affordable attention that produces nothing.
The hybrid formats worth knowing
The binary is a bit false in 2026. A few things sit in between and are frequently the best answer:
- Static-in-motion. A still image with subtle animation, a text reveal, or a slow zoom. Produced at near-static cost, counted as video by the platform, and often the best performance-per-dollar-of-production in the account.
- Carousels. Multiple statics in sequence. Good for multi-product, multi-benefit and step-by-step arguments, and they let one production session yield several angles.
- Screen recordings and demos. For software, the cheapest genuinely useful video there is. No shoot, no talent, high explanatory value.
- Creator-shot video. Cheaper than produced video, different in register, and the trade-offs are covered in UGC vs brand-produced creative.
If your account has a hard rule against one format, these are usually where the unclaimed performance is sitting.
The test that settles it for your account
Do not take a general answer. Run this:
- Pick one audience and one offer. Hold everything else constant.
- Build three statics and one video carrying the same core message. This ratio reflects real production economics, not a laboratory 1:1.
- Run them in a single ad set with enough budget to exit learning, and let the platform allocate. You are testing creative, not bidding.
- Judge on cost per outcome after a full purchase cycle, not on day-three engagement.
- Then check the account total. If the video won its comparison but account-level efficiency did not move, the video took budget from something that was already working.
Run it once a quarter rather than once. Creative answers expire — which is the subject of creative fatigue.
The short version
Video for cold audiences and for anything that needs explaining. Statics for warm audiences, retargeting, offers and objection handling. Statics to find the message, video to scale it. Judge both on cost per outcome at the account level, never on engagement metrics only one of them can produce.
And if your production budget is fixed — which it always is — remember that the format you can make fifteen of will usually teach you more than the format you can make two of.
If your creative is the constraint and you are not sure where the budget should go, that is the kind of question I work on with brands.